Psychoacoustic audio analysis in frequency domain

When evaluating the loudness sensation of an acoustical signal, the concept of energy summation in terms of the frequency components of the signal must be reformulated. There are 2 standardized methods for the calculation of the loudness of a signal. Both methods are based on the decomposition of the signal into frequency bands and consider an amplitude axis in Sones. As an approximation to those methodologies, a model is used, which conjugates the concepts of equal-loudness-contours and the Terhardt masking model for obtaining an amplitude weighting in Phons; meanwhile the energy summation is carried out using a Riemann sum approximation to integrate the specific loudness curve. Some results presented for this methodology are applied to audio signals and compared with the traditional one-third-octave band analysis.